3results about How to "Affect judgment" patented technology

This utility model discloses a chemotherapy pump quick calculation compass, relating to the field of medical device technology. The chemotherapy pump quick calculation compass includes a first turntable, a second turntable, a rotating pointer, and a rotating shaft. The first turntable, second turntable, and rotating pointer are sequentially arranged along the axial direction on the rotating shaft. The upper surface of the first turntable has a dosage scale, and its center is fixedly connected to the rotating shaft. The upper surface of the second turntable has a time scale, and its center is rotatably connected to the rotating shaft. The end of the rotating pointer furthest from the needle tip is rotatably connected to the rotating shaft. The diameter of the first turntable is larger than that of the second turntable, and the length of the rotating pointer is smaller than the radius of the second turntable. This utility model, through the combination of the pointer's value, the time scale, and the dosage scale data, can intuitively show the infusion start time, the medical staff shift change time, and the infusion volume. Then, by comparing the infusion volume on the turntable with the value on the infusion pump, it can be determined whether the infusion volume is within a reasonable range.

The application discloses a marker and a pushing device, and relates to the technical field of medical devices. The marker is a spring coil, and the spring coil is provided with a plurality of reflecting surfaces. The plurality of reflecting surfaces are divided into a dense reflecting part and a sparse reflecting part. The dense reflecting part is used for forming a dense bright band image during development, and the sparse reflecting part is used for forming a spaced bright band image during development. The application provides a marker for breast lesion positioning, which enhances the development of the marker under ultrasound, and the marker can quickly recover to an unfolded state after being released to a target position, and is firmly combined with tissue and is not prone to displacement.

The invention relates to the field of artificial intelligence and deep learning, in particular to a user action type recognition method based on artificial intelligence, which comprises the following steps: acquiring a user action video to be recognized; performing action analysis on the user action video, and selecting a plurality of continuous optimal video frames to generate an optimal video clip; performing human body target enhancement on the optimal video clip through a target enhancement model to generate a target enhanced video clip; generating a corresponding human body skin multi-person linear model based on the target enhanced video clip; and identifying the action type of the user in the to-be-identified user action video according to the human body skin multi-person linear model. According to the method and the device, the action fragment really reflecting the user action type is selected by performing action analysis on the user action video, and the user target in the video frame is better highlighted through human body target enhancement, so that the accuracy and the efficiency of identifying the user action type are improved.